Tutorial COFDM

8/3/2019 Tutorial COFDM

1/43

COFDM

A brief history

COFDM principles

COFDM transmission sequence

Countering against echoes and reflections

DVB-T framing structure

DVB-T variable parameters


2/43

COFDM

A brief history

COFDM principles






3/43

A brief history of COFDM - 1

Popular in the 1980s and used for digital audio broadcasting(DAB)

OFDM + QPSK modulation

OFDM - A special form of MCM Patent issued in the US in 1970 (number 3,488445) submitted by R.W.

Chang in 1966

Time domain signals used to ensure subcarrier orthogonality Major contribution by Shannon in defining waveforms in Euclidean space,

allowing definitions of orthogonality

No need for steep band pass filters

Sub-carrier spectra allowed to overlap

Need for real time FFTs

OFDM grew out of Multi Carrier Modulation (MCM) Military HF radio (late 1950s)

Divides stream into several parallel bit streams

Bit streams used to modulate several carriers


4/43

A brief history of COFDM - 2

DTG (UK based Digital Terrestrial Group, set up in 1995 to make a workingbroadcast solution for the UK to meet Government plans)

9 First commercial broadcasts in late 1998 with simulcast and laterOnDigital services.

Various associated bodies

1992DVB (Digital Video Broadcast - voluntary group of 200 companies)

9 DVB-S, DVB-C in 1994 and DVB-T in early 1997

dTTb (digtial Terrestrial Television broadcast project)

9 Demonstrator to show the feasibility of a commercial receiver

DVBird (Digital Video Broadcast integrated receiver decoder)9 Technical specifications needed and partitioning of electronic functions


5/43

COFDM

A brief history





COFDM principles


6/43

What is COFDM ?

C - Coded

O - Orthogonal

F - Frequency

D - Division

M - Multiplex


7/43

FDM (Frequency Division Multiplex)Data signals

Carriermodulation

fc1

Carriermodulation

fc2

Carriermodulation

fcn

:

s1(t)

s2(t)

sn(t)

Transimitter

Transmitter architecture

Output signal

G(t)

S1(f)

S2(f)

Sn(f)


8/43

FDM (Frequency Division Multiplex)

NRZ

1 1 0 0 0 1 1Code

ASK

FSK

PSK

- Modulation

Quadrature (QAM)phase shift uses a/2 phase shift. phase shift isshown here


9/43


Guard interval

Frequency

Time

Useful data

Continuous frequency

transmissions G(t)

FC1

FC2

FC3

FC4

FC5

6817 or 1705 frequenciesFCn

Tu

Tu 1/(Fc1 - Fc2)


10/43


ndata symbols over time period T

time

G(t)


11/43


Receiver architecture

BPFc1

BPF = Band pass filter

BPFc2

BPFcn

:

Carrier

demod

Carrier

demod

Carrier

demod

G(t)

s1(t)

sn(t)

s2(t)


12/43


IDFT

DFT

s(ti) =2WN

fk=1

N

S(fk)ej2fkti/N

S(fk) =TN

ti=1

N

s(ti)ej2fkti/N

Carriermodulation

fc1

Carriermodulation

fc2

Carriermodulation

fcn

:

s1(t)

s2(t)

sn(t)

Transimitter

Transmitter architecture

Output signal

G(t)

S1(f)

S2(f)

Sn(f)

Receiver architecture

BPFc1

BPF = Band pass filter

BPFc2

BPFcn

:

Carrier

demod

Carrier

demod

Carrier

demod

G(t)

s1(t)

sn(t)

s2(t)


13/43

COFDM

A brief history

COFDM principles






14/43

DVB-T framing structure Fixed number of carriers used

9 Allows receiver to lock onto signal

Modulation used

9 Increases number of bits that can be transmitted

9 '2K' system in UK (1705 carriers)

Carrier types

9 Keeps constant power levels

9 Eg each carrier transports 4 bits for QAM-16

9 '8K' also an option (6817 carriers)

9 Data carriers - 2,4 or 6 bits per symbol, per carrier

9 TPS carriers - Transmission information9 Pilot carriers -Channel estimation at receiver, Tx at boosted power levels

Scattered - 524 in '8K' mode, 121 in '2K' pseudo random within symbol

Continual - 177 in '8K' mode, 45 in '2K' always in same position within symbol


15/43


f

Single frequency carrier.One of 6817 (8k) or 1705 (2k)

discrete modulation carriers.

Either: Data (6048 or 1 512)Continual pilot (177 or 45)

Scattered pilot (524 0r 131)

TPS carrier (68 or 17)

f

....

6817 carriers (8K)

1705 carriers (2K)

OFDM symbol

(frequency domain)


16/43


OFDM symbol

(time domain)

....

OFDM frame

0 67

Tf

t

Useful data

Guard

interval

Tu

Ts

t

t

OFDM super frame

4 x OFDM frames


17/43

COFDM

A brief history

COFDM principles






18/43

COFDM functional blocksOuter coding*

(R/S bytes added)

204 204

RS

MPEG-2 Transport streaminput

Randomisation*

PRBS

188 188 188

Outer interleaving*

(Forney)

Inner coding*

|||||||........||||||||

Bit and symbol interleaving

....011001010001.... |||||||........||||||||

|||||||........||||||||

Amplitude/phase

mapping

R|||||||........||||||||

|||||||........|||||||| I

Pilots and TPS addition

Inverse FFT........................

|||||||........||||||||R

|||||||........||||||||IIFFT

FIR

Time shift andcombination

Guard interval

insertion

|||||......|||||||||||......||||||

GI

Analogue

conversion Upconversion

F

DAC

Filtering Transmission

* Same as DVB-S


19/43

DVB transport stream

TELETEXT_1

VIDEO_1

AUDIO_1

TELETEXT_2

VIDEO_2

AUDIO_2

TELETEXT_3

VIDEO_3

AUDIO_3

TELETEXT_4

VIDEO_4

AUDIO_4

TP1_1 TP1_2 TP1_3 TP2_1 TP2_2 TP2_3 TP3_1 TP3_2 TP3_3 TP4_1 TP4_2 TP4_3

ES

TP

PROGRAMS

TP1_1 TP2_1 TP1_2 TP4_1 TP3_1 TP2_2 TP2_3 TP1_3 TP4_2

TRANSPORT

MUX

TRANSPORTSTREAM

PCR_1 PCR_2 PCR_3 PCR_4


20/43

Data scramblingPseudo Random Binary Sequence (PRBS)

Energy dispersal to ensure adequate binary transitions

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Enable (1)

Data input

Randomized data

output

1 0 0 1 0 1 0 1 0 0 0 0 0 0 0Initialisation

sequence


21/43

Error correction Error prone environment hence small packets (188 bytes) with

additional error correction data (16 bytes)

Known as Forward Error Correction (FEC) Also known as channel coding

Two main parts:

Outer coding for burst errors (Reed - Solomon and Forney)

Inner coding (Convolution coding)

QEF Channel

Transmitter

FEC

Receiver

FECData Data

Energy

dispersal

Outer

coding

(RS)

Forney

Interleaving

Inner

coding

(Convolution)

BER < 10-10


22/43

Outer coding

Reed Solomon

Operates over individual packets Corrects up to 8 erroneous bytes per packet

Non correctable flag for > 8 byte errors

Bandwidth overhead is 8%

Forney convolution interleaving Increases efficiency of the RS coding

Spreads errors over a greater area


23/43

Inner coding

Convolution coding

2 identical streams produced from outer coded stream Output stream formed from combination of these new streams

Not all simultaneous bits taken - hence rate defined

(DVB-T code rates: 1/2, 2/3, 3/4, 5/6, 7/8)

Puncture rate impact on data rate Puncture rate of 3/4 means 1 out of 4 bits is removed

Data rate becomes: (1/2)*(4/3) = 2/3 of original (ie code rate is 2/3)

No puncturing data rateis halved since convolutionencoder produces twoidentical streams

Every 4th bit removed


24/43

Bit and Symbol Interleaving

Bite-wise interleaving

Inner coder has two output streams Bit wise interleaver produces 2, 4 or 6 streams for QPSK, 16-QAM and

64-QAM respectively

Symbol interleaving The 2, 4 or 6 bit words are mapped onto the OFDM carriers

1512 for 2k mode or 6048 for 8K mode


25/43

Amplitude and Phase Mapping

(example)

Q

I

12 phases / 3 amplitudes

2 amplitudes appear on 4 phases

1 amplitude appears on 8 phases

16-QAM


26/43

Pilots and TPS addition Pilots

Continual pilots

9Always in the same place within the OFDM symbol

9 45 in 2k mode, 177 in 8k mode

9 Transmitted at increased power levels

9 Used to estimate the channel characteristics and therefore make corrections

Scattered pilots

9 Located as a pre-defined pattern such that there is an equal number persymbol

9 131 in 2K mode, 524 in 8k mode

9 Transmitted at increased power levels

9Used in conjunction with continual pilots to estimate the channel distortion


27/43

Pilots and TPS addition TPS (Transmission Parameter Signalling)

Type of modulation used

Hierarchy information

Guard interval

Inner code rates

Transmission mode (ie 2k or 8k)

Frame number within a super frame (ie 0 to 3)

DPSK (Differential Phase Shift Keying) modulation used due to robustness


28/43

IFFT, time shift and combination

Complex to real conversion

Q (real) and I (Imaginary) are added, sampled and output

IDFT at transmitter, DFT at receiver FFT actually used (computational algorithm) for summing operation

FFTs must be powers of 2, hence 2k or 8k modes

Much faster that normal DFT

Eg if 8k point DFT takes 670 ms then the FFT takes .53 ms

N

2 log NN2


29/43

Guard Interval Insertion Replication of end of symbol placed at beginning

Useful symbol

t

Main signal

Copy of end of symbol

Also means receiver canidentify start of symbol usinga correlation function


30/43

Final stages - Transmission D/A conversion Filtering

Upconversion and transmission

FM sound carrier

NICAM

COFDM carrier

530 MHz +/- 1/6 Mhz

(Centre)

526 MHz

f

CH27

519.25 MHz518 MHz

CH28

Video luminance

carrier

Video picture

information

7.61 MHz

Analogue

transmission

Digital

transmission

Example: UK transmission in channel 28


31/43

COFDM functional blocksOuter coding*

(R/S bytes added)

204 204

RSMPEG-2 Transport streaminput Randomisation*

PRBS

188 188 188

Outer interleaving*

(Forney)

Inner coding*

|||||||........||||||||

Bit and symbol interleaving

....011001010001.... |||||||........||||||||

|||||||........||||||||

Amplitude/phase

mapping

R|||||||........||||||||

|||||||........|||||||| I

Pilots and TPS addition

Inverse FFT........................

|||||||........||||||||R

|||||||........||||||||IIFFT

FIR

Time shift andcombination

Guard intervalinsertion

|||||......|||||||||||......||||||

GI

Analogue

conversion Upconversion

F

DAC

Filtering Transmission

* Same as DVB-S


32/43

What is COFDM ?

C - Coded

O - Orthogonal

F - Frequency

D - Division

M - Multiplex


33/43

Orthogonality Spacing between carriers is minimised

Results close to theoretical maximum are achieved (f 1/)

1/T

fk fk+1 fk+2 fk+3 fk+4

Expensive in analogue FDM due to costly band pass filters

Definition possible due to signals being described as vectors


34/43

COFDM

A brief history

COFDM principles





C t i i t h d


35/43

Countering against echoes and

reflections

Repetition of signal to counter echoes Echoes caused by

Moving receiver

Moving transmitter

Reflection from moving or static objects

Set top boxBuilding

COFDM transmitter

Direct signal

Reflected

signal

Set top box

COFDM transmitter 1

Direct signal

COFDM transmitter 2

Distant signal

Single Frequency Networks (SFNs)

C t i i t h d


36/43

Countering against echoes and

reflections

Echo length is easily calculated

Assuming 2k Mode with Guard interval 1/32 1/32 of the symbol transmits in 7 us

Maximum delay = 7 us

Distance = 3 x 108 m/s x 7 us

Distance = 2.1 km

Useful data

Guardinterval

t

Main signal

Useful data

Guard

interval

t

Delayed signal

Max. delay


37/43

COFDM

A brief history

COFDM principles






38/43


Carrier mode: 2k or 8k

Type of modulation: QPSK, 16-QAM, 64-QAM Guard Interval: 1/4, 1/8, 1/16, 1/64

Inner code rate: 1/2, 2/3, 3/4, 5/6, 7/8

Hierarchical modes

Selection of transmission bandwidth (6/7/8 MHz)


39/43


Modulation

QPSK

16-QAM

64-QAM

Code rateGuard interval

1/2

2/3

3/4

5/6

7/8

1/2

2/3

3/4

5/6

7/8

1/2

2/3

3/4

5/6

7/8

1/4 1/8 1/16 1/32

4.98

6.64

7.46

8.29

8.71

9.95

13.27

14.93

16.59

17.42

14.93

19.91

22.39

24.88

26.13

5.53

7.37

8.29

9.22

9.68

11.06

14.75

16.59

18.43

19.35

16.59

22.12

24.88

27.65

29.03

5.85

7.81

8.78

9.76

10.25

11.71

15.61

17.56

19.52

20.49

17.56

23.42

26.35

29.27

30.74

6.03

8.04

9.05

10.05

10.56

12.06

16.09

18.10

20.11

21.11

18.10

24.13

27.14

30.16

31.67

Useful data rate (M bits / sec)


40/43


Significance of mode and guard interval

8k system allows good reception with long multi-path echoes 8k system is therefore suitable for single frequency networks (SFNs)

2k system more suited to multi frequency or single transmitter networks

A larger guard interval implies a lower bit-rate efficiency

The guard interval value is therefore a trade-off between bit-rate and

network tolerance to echoes and reflections


41/43

DVB-T hierarchy codingI

Q

64 QAM constellation

Low priority carriers


42/43

DVB-T hierarchy codingI

Q

In poor S/N ratio conditions16 64-QAM constellation pointscan be demodulated as one QPSKconstellation point

High priority carriers


43/43

DVB-T hierarchy coding Transmission of the same or different data for:

Same or different program can be transmitted in HD and SD + greater

error recovery

Poor reception areas can view SD if HD not possible

Transmission of different resolutions / characteristics:

Reception by different cost receivers (high end, low end, mobile,portable)

Other data can be transmitted related to the program

Date post:	06-Apr-2018
Category:	Documents
Upload:	noor-alamar
View:	248 times
Download:	0 times

Tutorial COFDM

Documents